Evaluation of the processes affecting vertical water chemistry in an alluvial aquifer of Mankyeong Watershed, Korea, using multivariate statistical analyses

Byoung Young Choi, Hyeon Jung Kim, Kangjoo Kim, Seok Hwi Kim, Hwa Jin Jeong, Eungyu Park, Seong Taek Yun

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

Vertical variations of redox chemistry and groundwater quality were investigated in an alluvial aquifer beneath an agricultural area, in which deep groundwaters are free of NO3, Fe, and Mn problems that are frequently encountered during the development of alluvial groundwaters. This study was performed to identify and evaluate vertical chemical processes attenuating these chemical species in the study area. For this study, the processes affecting groundwater chemistry were identified by factor analysis (FA) and the groundwater samples collected from six multilevel samplers were hierarchically classified into three different redox zones by cluster analysis (CA) based on the similarity of geochemical features. FA results indicated three major factors affecting the overall water chemistry: agricultural activities (factor 1), redox reactions (factor 2), and remnant seawater (factor 3). The groundwater quality in the study area was revealed to be controlled by a series of different redox reactions, resulting in different redox zones as a function of depth. It was also revealed that the low Fe and Mn levels in the groundwater of the deeper part are associated with sulfate reduction, which led to precipitation of Fe as iron sulfide and adsorption of Mn on it.

Original languageEnglish
Pages (from-to)335-345
Number of pages11
JournalEnvironmental Geology
Volume54
Issue number2
DOIs
StatePublished - Mar 2008

Keywords

  • Alluvial groundwater
  • Cluster analysis
  • Factor analysis
  • Mankyeong watershed
  • Redox reaction

Fingerprint

Dive into the research topics of 'Evaluation of the processes affecting vertical water chemistry in an alluvial aquifer of Mankyeong Watershed, Korea, using multivariate statistical analyses'. Together they form a unique fingerprint.

Cite this